外文翻译---宽带接入技术

附录1 外文文献译文宽带接入技术宽带接入技术是在电信和视频网络行业的催生影响下产生出来的，其应用主要体现在物理层。

相反，数据网络行业的关注点主要集中在网络层和运输层上（例如IP电话）。

形成这种应用方法上的区别的理由是，一旦传输的内容被数字化，就能作为数据进行处理。

换句话说，从网络的观点来看，网络只是以数字的形式来传输数据。

在应用层上，这些数据位不论实际代表话音、数据，还是代表视频流都没有区别。

因此，数据网络行业在让数据流适应跨网传输这方面并没有做什么。

另外一方面，电信和视频网络行业必须解决利用现有接入基础设施将内容以分组数据传输的方法进行传输。

这成为在开发下述几类宽带接入技术中的重要催化剂。

铜回路接入技术，又称为数字用户线（DSL）技术。

DSL技术又统称XDSL。

XDSL是由电信行业开发的，利用了世界范围内的几百万英里长的现有铜回路通信设施。

光纤/同轴电缆的电缆接入技术。

电缆接入是由视频网络行业中的有线电视部分开发的，利用了已经连接全世界几百万个（大多数是常驻的）用户的视频信道电缆基础设施。

在线卫星接入技术。

卫星接入是由视频网络行业中的无线部门开发的，通过使用卫星基础设施，经视频信道接到有线电视中心终端局并传送到世界范围内的几百用户（大多数是常驻的）。

非对称数字用户线（ADSL）20世纪80年代，ADSL作为电信行业为应对电缆行业支持视频点播需求而提供的一种解决方案，首先被开发了出来。

然而直到20世纪90年代中期才认识到，它也可以作为能访问像因特网这样的高速业务的技术。

ADSL提供非对称传输，典型的下传速率可达到9Mb/s (从中心局到设备），上传速率为16Kb/s到640Kb/s （从设备到中心局）。

像所有的铜线传输系统一样，速率越高，传输范围越短。

混合光纤同轴网技术（HFC）混合光纤同轴网技术是基于现有的有线电视（有线TV或CATV）技术的。

最初，有线电视是基于从多系统运作员（MSO）到客户设备的同轴电缆设施的，并采用树形拓扑。

这些系统大多数已升级到HFC。

在这种系统中，信号通过一对光缆被传送到一个光缆节点上，然后再经过同轴电缆分发给客户。

在中心终端局上，各种源信号，如传统的卫星业务、使用广域网的模拟和数字业务，以及使用专用主干网的因特网服务提供商（ISP）的业务被复用，并从电信号射频转换成为光信号。

光缆上的通信是单向的：从中心终端局到光缆节点的每一对光缆，都以不同的方向承载单向业务。

光信号在光缆节点处转换成射频信号并以双工方式沿同轴电缆传送。

从中心终端局到客户设备的信号称为下传信号或者前向通路信号。

从客户设备到中心终端局的信号称为上传信号或者反向通路信号。

光纤通信介绍80年代一项最最重要的技术发展是光纤通信成为一个主要的国际性产业。

用光纤敷设总长度可以表明其发展程序。

据估计，到1987年底仅美国的光纤敷设总长将达320万公里，其中90% 以上是在1982-1987年间敷设并开通的，而长度干线占主导地位，数量约为光纤总长的95% 。

虽然现在人们对于纤维光学的兴趣主要在于通信，但早期发展纤维光学的目的并不在此。

50年代初研究人员制造出第一根具有包层玻璃光纤时，并不想用于通信而是想用它们传送内窥镜需要的成像光束。

1966年Kao和Hockham发表了那篇著名的论文，建议将低损耗光纤用于通信，此时纤维光学已发展为一项很实用的技术了。

1970年10月，第一根低损耗（20dB/km）石英光纤问世了。

有时将这一日期作为光纤通信时代的开端。

虽然这一成果当时在研究领域确实引起了极大的关注，但这种光纤距离通信所要求的条件还相差甚远：每千米20dB的损耗对于长途通信系统仍然是太大了；光纤易断裂，必须寻找保护方法；没有合适的光源。

研究人员不知道光缆的终端和接头是否会发展到实用阶段，至于生产这些器件是否在经济上可行，从而使之在市场上占有重要地位，他们更是存有严重的疑虑。

虽然技术障碍好像不可逾越，但经济潜力却非常明显。

正因为如此，在70年代早期研究和开发工作发展迅速，一些重要问题得以解决。

70年代中后期，由于发展重点由研究领域转入工程实用，因而加速发展了适应市场的产品。

在实验室研制的光纤损减值接近瑞利极限值：0.8µm波长处为2dB/km，1.3µm波长处为0.3dB/km和1.55µm波长处为0.15dB/km。

通过改进光纤外涂层方法和成缆技术，克服了微弯损耗。

生产了加强型光和多纤连接器用于室外作业。

工作在0.8-0.85µm波长区的商用镓铝砷激光器的室温阈值电流减少到20-30mA范围。

据称，激光器和发光管的设计寿命达10万到100万小时。

开发了工作于1.3µm波长附近的光源和改进的光电检测器，从而可以利用光纤在长波长区的低损耗和低色散特性。

这一时期进行的室外实验较重要的有AT&T于1976年在亚特兰大的实验，1977年在芝加哥的实验和1977年日本的用户通路项目。

到了80年代，光纤器件在性能、价格和可靠性方面的改善使众多电话公司受益匪浅。

光纤很快成为长途干线的首选传输媒质。

一些早期敷设的光缆线路采用0.8µm光源和渐变折射率多模光纤，但到1983年，城市间线路的设计者们就考虑使用1.3µm单模光纤系统了。

单模光纤与1.3µm激光器相连，可以提供宽带特性，增加了高速率系统的中继距离。

最近敷设的光纤系统的数据速率已移至每秒几比特范围。

这种系统采用光谱纯的分布反馈激光器，将光纤色散效应减至最小。

在1.55µm波长上设计的低色散光纤，相应地具有低损耗特性，目前广泛用于长途通信。

为进一步增加光纤的信息容量，逐渐广泛采用波分复用方法。

人们对于光纤在其他领域的潜力刚刚开始认识。

用于计算机系统和办公室的光纤网络逐渐变得更加重要。

在电话系统中，光纤在主要城市地区中心交换局间互联和低级交换中的使用继续迅速增加。

入户光缆已经有了示范工程。

许多观察家相信，全国电话系统将使用光纤传输视频宽带信号而逐渐升级。

这些宽带用户环路系统将为可视电话、视频娱乐节目等业务提供通路。

宽带业务广泛使用光纤将会变得经济可行。

异步转移模式ATM（异步转移模式）既是复用技术又是交换技术。

最初，人们是想用ATM 来处理高比特率的数字信号，事实却证明它是一种通用技术，可以用来传输和交换任何类型并具有各种比特率的数字化信息。

无论传输的信息是什么，ATM都以称作“ 信元”的短的分组采传送信息。

信元是由固定的48字节加上5字节的信头组成。

信元寻找路由是基于带有双重识别的逻辑信道。

ATM既与电路方式有关又与分组方式有关。

由于使用简单的协议，信元至网络节点的转移可完全由硬件处理完成，这就缩短了转送时间，提高了传输路径的速率，使比特速率甚至可以达到每秒几百兆比特。

另一方面，ATM保留了分组方式所有的灵活性：只传送所需要的信息，提供简单、独特的复用方法而不管不同信息流的比特率，并且允许比特流的变化。

ATM开始于80年代初，那时人们试图找到一种更适合的技术用于交换超过100Mbit/s的高比特率的信道。

1988年，ITU批准了I .121建议，该建议选择ATM 作为用于各种类型信息宽带网络的目标传送模式，其中包括如话音的低比特率的信息。

1991年一些运营公司和厂商建立了ATM论坛以加速ATM标准化工作。

现在ATM论坛的成员已超过600个，它对A TM 的标准化和规范化有着重要影响。

第一批ATM产品1992年面世，主要用于局域网，其设计主要是用来解决计算机终端间随计算机数量和功能不断增加所带来的共享同一承载电路的问题。

ATM网络可以近似在看作是由三个覆盖功能层组成：业务和应用层、ATM 网络层和传输层。

应用层提供端到端的业务。

应用层使用ATM网络层的逻辑连接，当信元通过由逻辑连接（称作虚连接）共享的传输链路时，ATM网络层依次对信息流复用并寻找信息流的逻辑路由。

传输层提供物理链路并处理信元的实际物理传输。

ATM网络能够传输和交换话音、数据和视频业务，从接入的角度看这些业务使用传统的数字接口并具有同样的服务质量。

这就意味着任何两个终端间的物理连接都可由等效的逻辑连接代替，逻辑连接可在共用的传输。

链路中与其他的逻辑连接复用。

资源可在所有连接中动态共享与同步时分复用技术相比，同步复用技术僵硬地将业务与传输资源相连，而异步技术的优势是根据其确切的需要来占用传输链路。

ATM技术将网络传输的应用和业务与所使用的传输资源完全分开。

构成虚网络的能力意味着物理网络可以由许多用户动态实时地共享，因而使网络结构得到低价高效的使用，对高比特率业务也一样。

对所有网络层的投资都是适应未来需要的，因为不同的应用在出现新的需求时可及时在同一网络结构中进行重新分配。

ATM提供一种独特的方式将传输不同业务的网络协调成单一的物理网络。

以上所描述的ATM优点说明了为什么人们对ATM充满热情。

随着数字化和图像编码技术的进步，交互视频业务和更通常的多媒体业务开始出现。

这些业务将会对网络产生很大的影响。

今天，ATM是唯一能够提供这些业务所需的高比特率和灵活性的传输的技术。

ATM，远比任何其他电信技术更能满足运营公司和用户对当前和未来业务的需求。

与其他有可能在某些应用领域与ATM竞争的技术相比，ATM（主要是由于其通用性，无论是比特率还是传输的信息类型）都具有特殊的优点。

ATM 对所有比特率的信号都可提供交换功能，这一点特别适合于高比特率和可变化比特率信号。

ATM的独特性将使它成为卓越的多媒体业务的自然载体，特别是对于可变比特率的视频，并且使它成为能够提供如视频点播新业务的未来信息高速公路必不可少的一部分。

在很短的时间内，运营者对ATM产生了很大的兴趣，主要是由于将连接的概念与实际资源分开所引入网络的灵活性和虚拟性。

这就简化了网络的管理功能并能最佳地使用网络资源，特别是通过统计复用和建立虚拟专用网络。

当然，在ATM技术普遍使用之前仍会有很长的路要走，但是这场正在进行的技术革命将会深刻在影响数据处理和视频处理，影响电信世界。

这一革命所产生的影响无疑会比在模拟网络中出现数字技术的影响要大得多。

附录2 外文文献原文BROADBAND ACCESS TECHNOLOGIES Broadband access technologies have been spawned by the efforts of the telecommunications and video networking industries to move toward convergence at the physical layer. By contrast, the efforts of the data networking industry toward convergence have generally focused on the networking and transport layers (e.g., IP telephony). The reason for this difference in approach is that, once content has been digitized, it is treated as data. In other words，from a networking perspective, the network merely transports data in digital format. It makes no difference whether the bits actually represent voice，data，or video steams at the application level. Consequently, the data networking industry did not have to do anything different to adapt a data stream for transport across a network. On the other hand, the telecommunications and video networking industries had to figure out way to transport their content as packetized data cost-effectively by working with existing access infrastructure. This was an important catalyst in the development of the following classes of broadband access technologies.Copper-loop access technologies, also known as digital subscriber line (DSL) technologies. DSL technologies are collectively referred to as XDSL.XDSL was developed by the telecommunications industry to make use of the several million miles of existing copper loop infrastructure around the world.Cable access technologies over fiber/coaxial cable. Cable access was developed by the cable TV portion of the video networking industry to take advantage of the cable infrastructure that feeds video channels to several million (mostly residential) subscribers around the world.Satellite access technologies over wireless medium. Satellite access was developed by the wireless portion of the video networking industry to make use of the satellite infrastructure that feeds video channels to cable TV head end offices and to several hundred(mostly residential) subscribers around the world.ADSLADSL was first developed in the 1980s as the telecommunications industry's answer to the cable industry' request to support video on demand. In the mid 1990s, however, it was quickly recognized as a viable technology to enable access tohigh-speed services such as the Internet. ADSL delivers asymmetric transmission rates typically up to 9 Mbps downstream (from the CO to the premises) and 16 Kbps to 640 Kbps upstream (from the premises to the CO). Like all copper transmission systems, the higher the bit rate, the shorter the range.HFC TechnologyHFC technology is based on existing cable television (cable TV or CATV) technology. Originally, cable TV systems were based on coaxial cable facilities from an MSO to a customer premises and used a tree topology. Most of these systems have been upgraded to HFC, by which the signal is brought to a fiber node via a pair of optical fibers and then distributed via coaxial cable to customers. At the head end, signals from various sources, such traditional satellite services, analog and digital services using WAN, and Internet service provider (ISP) services using private backbone network, are multiplexed and converted up from an electrical (radio frequency) signal to an optical signal. Communication is one way on the optical fiber; each of a pair of optical fibers from the head end to the fiber node carries one-way traffic in opposite directions. The optical signal is converted down to RF at the fiber node and travel over the coaxial cable in duplex mode. The signal going from the head to the customer premises is called a downstream signal or a forward path signal. The signal going from the customer premises to the head end is called an upstream signal or a reverse path signal.Introduce to Optical Fiber CommunicationOne of the most important technological developments during the 1980s has been the emergence of optical fiber communication as a major international industry. One indication of the extent of this development is the total length of installed fiber, which was estimated to be 3.2 million kilometers in the U.S. alone by the end of 1987. Over 90% of this fiber was placed in service during the time period of 1982~1987. Long-haul trunk installations have been dominated, accounting for about 95% of the fiber in the U.S.Although telecommunication is the rationale for most of the current interest in fiber optics, this was not the case during the early days of the technology. The researchers who produced the first clad glass optical fibers in the early 1950s were not thinking of using them for communications; they wanted to make imaging bundles for endoscopy. Fiber optics was already a well-established commercial technology when the famous paper by Kao and Hockham, suggesting the use of low-loss optical fibersfor communication, appeared in 1966.The first low-loss (20dB/km) silica fiber was described in a publication which appeared in October of 1970. The date of this publication is sometimes cited as the beginning of the era of fiber communication. Although this development did receive considerable attention in the research community at the time, it was from inevitable that a major industry would evolve. The 20dB/km loss figure was still too high for long-hall telecommunication systems. The fibers were fragile, and a way to protect them would have to be found. There were no suitable light sources. Researchers did not know whether field termination and splicing of optical cables would ever be practical. Finally, there were serous doubts as to whether these components could ever be produced economically enough for the technology to play a major role in the marketplace.Although the technological barriers appeared formidable, the economic potential was very significant. As a consequence, research and development activity expanded rapidly, and a number of important issues were resolved during the early 1970s.During the middle and late 1970s, the rate of progress toward marketable products accelerated as the emphasis shifted from research to engineering. Fibers with losses approaching the Rayleigh limit of 2dB/km at a wavelength of 0.8um, 0.3dB/km at 1.3um, and 0.15dB/km at 1.55um, were produced in the laboratory .Microbend loss problems were overcome through the use of improved fiber coatings and cabling techniques . Rugged cables and multiform connectors were produced for field installation. Rom temperature threshold currents for commercial gallium aluminum arsenide lasers operating in the 0.8 to 0.85um spectral region were reduced to the 20 to 30 mA range, and projected lifetimes in the 100,000 to 1,000,000 hour range were claimed for both lasers and LEDs. Light sources and improved photodetectors which operated near 1.3um, were developed to take advantage of the low fiber loss and dispersion in this “longer wavelength region”. Several major field trials were undertaken during this period, including AT&T，s Atlanta experiment and Chicago installation, and Japan，s subscriber access project.Improvements in component performance, cost, and reliability by 1980 led to major commitments on the part of telephone companies. Fiber soon became the preferred transmission medium for long-haul trunks. Some early installations used 1.8um light sources and graded-index multimode fiber, but by 1983, designers of intercity links were thinking in terms of 1.3um, single-mode systems. Thesingle-mode fiber, used in conjunction with a 1.3um laser, provides a bandwidth advantage which translates into increased repeater spacings for high data rate systems.Data rates for installed fiber optic systems have recently moved into gigabit per second range. Such systems use the spectrally pure distributed-feedback lasers to minimize fiber dispersion effects. Fibers designed for low dispersion at 1.55um wavelength, which corresponds to minimum fiber lass, are now commonly used in long distance transmission. The use of wavelength multiplexing to further increase the fiber information capacity is becoming more widespread.The potential of fiber optics in other areas is only beginning to be realized. Fiber optic networks for computer systems offices are becoming more prominent. In the telephone system, the use of fiber optics for interconnecting central offices within a metropolitan area and for lower levels in the switching hierarchy is still increasing rapidly. Fiber links to the home have been used in demonstration projects. Many observers believe that national telephone systems will eventually be upgraded to handle video bandwidths by using fiber optics. These wideband subscriber loop systems would provide access to services such as picturephone, video entertainment. Widespread installation of these broadband services will become economically feasible.ATMATM (Asynchronous Transfer Mode) is both a multiplexing and switching technique. It was initially intended to handle high bit rates, but it has in fact proved to be a universal technique for transporting and switching any type of digitized information at a wide variety of bit rates.ATM transfers information in short packets called “cells” with a fixed length of 48 bytes plus five header bytes, irrespective of the underlying type of transmission. Cell routing is based on the principle of logical channels with dual identification: the cell header contains the identifier of the basic connection to which the cell belongs-called a virtual circuit (VC) and the identifier of the group of VCs to which the connection belongs-called a virtual path (VP) .ATM is related to both circuit and packet modes. Because of the simplicity of the protocol used, the transfer of cells to the network nodes can be handled entirely by hardware, which leads to very short transit time and high usage of transmission paths, even at bit rates of several hundred megabits a second. On the other hand, ATM retains all the flexibility of the packet mode, enabling only required information to beconveyed, offering a simple, unique multiplexing method irrespective of the bit rates of the different information flows, and allowing these bit rates to be varied.ATM dated from the beginning of the 1980s: at the time , people were trying to find the most suitable technique for switching high bit rate channel at more than 1000 Mbit/s . In 1988, the ITU approved recommendation 1.121 which ratified the choice of ATM as the target transfer mode for broadband networks for all types of information, including low bit rate information such as voice. In 1991 several operators and manufactures founded the ATM Forum to expedite standardization. The ATM Forum now has more than 600 members and has a significant influence on ATM standards and specifications. The first ATM products appeared on the market in 1992: they were for local area networks and were designed to solve the problems of sharing the same bearer circuit between computer terminals as they continue to increase in numbers and power.An ATM network can be considered, in a first approximation, as being three overlaid functional levels: a service and applications level, an ATM network level and a transmission level. The applications provide an end-to-end service. They use the logical connections of ATM network level which in turn multiplexes and logically routes the information flow as ATM cells go through the transmission links shared by logical connections called virtual connections.The transmission level provides these physical links and handles the actual physical transport of the cells.An ATM network can transport and switch voice, data and video which, seen from the access, use traditional digital interfaces with the same quality of service. This means that a physical connection between any two terminals can be replaced with an equivalent logical connection which is multiplexed with others in a common transmission link. The resource is shared dynamically between all the connections.Compared with the synchronous time division multiplexing techniques which rigidly link service to resource, the asynchronous technique has the advantage of occupying the transmission link only in proportion to the exact requirement.The ATM technique completely separates the applications and services transported over a network from the transmission resources used. The ability to construct virtual networks means that the physical network can be shared by many users dynamically and in real time , thereby achieving cost-effective use of infrastructure, for high bit rate services too. Investments at all levels are also future-proofed, because of thedifferent applications can be reallocated in time over the same network infrastructure as requirements arise. ATM offers a unique way of coordinating different networks carrying different services into a single physical network .The advantages described above have explained the enthusiasm for ATM.As digitization and image encoding progress, interactive video services, and more generally multimedia services, are starting to emerge. Their impact on the network will be considerable. Today, ATM is the only transfer technique to offer the high bit rates and flexibility required by these services.ATM, much more than any other telecommunications technique, is able to meet the current and the future requirements of both operators and users. Compared with other techniques that may compete in certain applications, ATM is special mainly due to its universal nature, both in terms of bit rate and type of information transferred. ATM offers a switching function for all bit rates and this is particularly suitable for high and variable bit rates.ATM’s specific features will make it the preeminent nature vehicle for multime dia services, and especially for varying bit rate video, and will make it one of the essential components of future information superhighways offering new services such as video on demand. In the short term, ATM is also proving of great interests to the operators, because of the flexibility and virtuality that it can introduce into networks, by separating the concept of connection from that of physical resources. This simplifies network management functions and makes optimum use of resources, particularly through statistical multiplexing and the creation of virtual private networks.Of course, there is still a long way to go before the ATM techniques is in general use , but a revolution is underway which will deeply affect the worlds of telecommunications, data processing and video. The impact of this upheaval will without any doubt be greater than the advent of digital techniques in analogue networks.。